Electrolyte recommendation method and system, storage medium and program product

Abstract

The invention discloses an electrolyte recommendation method and system, a storage medium and a program product, relates to the technical field of lithium batteries, and recommends an electrolyte formula meeting a preset requirement through a pre-trained machine learning model based on multiple electrolyte formulas, multiple electrolyte test parameters and multiple electrolyte formula theoretical parameters. According to the method, the incidence relation between the electrolyte formula and the test parameters and / or theoretical parameters is learned by the pre-trained machine learning model, and the electrolyte formula meeting the preset requirement is recommended according to the incidence relation, so that a tester does not need to test a large number of different formulas to develop the electrolyte formula, the efficiency is relatively high, and the cost is relatively low. And the problems of poor objectivity and inaccuracy caused by complete dependence on manual subjective understanding and experience analysis are avoided, and a new method capable of improving the development efficiency and the development accuracy of the electrolyte formula is provided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of lithium batteries, in particular to an electrolyte recommendation method and system, a storage medium and a program product. BACKGROUND

[0002] Lithium batteries have become a research hotspot in the energy storage field due to their high energy density. Electrolyte, as the "blood" of lithium batteries, plays a crucial role. Currently, the development of electrolyte formula mainly relies on manual experiments, which requires the configuration and testing of a large number of different formulas, and the material testing and characterization cycle is long, resulting in low efficiency. Moreover, experimental data mainly depend on the analysis of experimental personnel, which is highly subjective, and the accuracy of the developed electrolyte formula is not high. Therefore, there is an urgent need for a new method that can improve the efficiency and accuracy of electrolyte formula development. SUMMARY

[0003] The present application provides an electrolyte recommendation method, system, storage medium and program product, which aims to provide a new method that can improve the efficiency and accuracy of electrolyte formula development.

[0004] To achieve the above-mentioned purpose, the present application provides an electrolyte recommendation method, which comprises:

[0005] A plurality of electrolytes are configured based on a plurality of electrolyte formulas, the electrolyte formula comprising a plurality of chemical components for preparing the electrolyte and the proportion of the plurality of chemical components; the test parameters of each electrolyte and the theoretical parameters of each electrolyte formula are obtained respectively, the test parameters being parameters obtained by testing the electrolyte and characterizing the performance of the electrolyte, and the theoretical parameters being parameters obtained by theoretically calculating the electrolyte formula and characterizing the performance of the electrolyte; based on the plurality of electrolyte formulas, the test parameters of the plurality of electrolytes, and the theoretical parameters of the plurality of electrolyte formulas, an electrolyte formula meeting a preset requirement is recommended through a pre-trained machine learning model; wherein the machine learning model comprises an association relationship, the association relationship characterizing the relationship between the electrolyte formula and the test parameters and / or the theoretical parameters, and the preset requirement comprising that the test parameters and / or the theoretical parameters are within a preset range.

[0006] In an embodiment, the pre-trained machine learning model is obtained by collecting sample data, the sample data including a plurality of sample electrolyte formulations, sample test parameters corresponding to the plurality of sample electrolyte formulations, and sample theoretical parameters; training a machine learning model using a portion of the sample data to obtain the correlation; evaluating the performance of the machine learning model using another portion of the sample data that does not participate in the training; and optimizing the machine learning model according to the evaluation result of the performance of the machine learning model until the performance of the machine learning model meets a performance threshold.

[0007] The embodiment provides a specific implementation of training a machine learning model.

[0008] In an embodiment, the test parameters include physical property parameters representing physical properties of the electrolyte and electrochemical parameters representing electrochemical properties of the electrolyte; and the test parameters of each electrolyte and the theoretical parameters of each electrolyte formulation are obtained by: performing physical property tests and electrochemical tests on each electrolyte to obtain the physical property parameters and the electrochemical parameters of each electrolyte; and performing theoretical calculations on each electrolyte formulation based on first principles and / or molecular dynamics simulation to obtain the theoretical parameters of each electrolyte formulation.

[0009] In an embodiment, the physical property tests include one or more of viscosity tests of the electrolyte at different temperatures, functional group tests of the electrolyte, solvation structure tests of the electrolyte, saturation gas pressure value tests of the electrolyte at different temperatures, and conductivity tests of the electrolyte; the electrochemical tests include one or more of Coulomb efficiency tests, limiting current density tests, electrochemical window value tests, ion diffusion coefficient tests, and alternating current impedance tests; and the theoretical parameters include one or more of density, ion binding energy, molecular dissociation energy, molecular surface electrostatic potential, dielectric constant, solvent donor number, solvent acceptor number, ion diffusion coefficient, and space charge density.

[0010] In addition, to achieve the above object, the application further provides an electrolyte recommendation system, which comprises a control module, an electrolyte configuration module and an electrolyte test module connected with the control module, and a computing device connected with the electrolyte test module.

[0011] The control module is configured to control the electrolyte configuration module to configure a plurality of electrolytes based on a plurality of electrolyte formulations, the electrolyte formulations including a plurality of chemical components for preparing the electrolytes and proportions of the plurality of chemical components.

[0012] The control module is further configured to control the electrolyte testing module to test each of the configured electrolyte to obtain a test parameter of each of the electrolyte, the test parameter being a parameter obtained by testing the electrolyte and representing a performance of the electrolyte;

[0013] The computing device is configured to obtain the test parameter of each of the electrolyte, and obtain a theoretical parameter of each of the electrolyte formula, the theoretical parameter being a parameter obtained by theoretically calculating the electrolyte formula and representing a performance of the electrolyte;

[0014] The control module is further configured to recommend an electrolyte formula meeting a preset requirement based on the plurality of electrolyte formulas, the test parameters of the plurality of electrolyte, and the theoretical parameters of the plurality of electrolyte formulas, by a pre-trained machine learning model, wherein the machine learning model comprises a correlation relationship representing a relationship between the electrolyte formula and the test parameter and / or the theoretical parameter, and the preset requirement comprises that the test parameter and / or the theoretical parameter is within a preset range.

[0015] In an embodiment, the electrolyte configuration module comprises a first mechanical arm, a reagent bottle, a powder injection device, a liquid injection device, a weighing scale, a heating and mixing device, and a visual recognition device.

[0016] The control module is configured to control the first mechanical arm to take the reagent bottle and place the reagent bottle on the weighing scale.

[0017] The powder injection device is configured to inject a powder into the reagent bottle on the weighing scale based on one of the plurality of electrolyte formulas.

[0018] The liquid injection device is configured to inject a solvent into the reagent bottle on the weighing scale based on the electrolyte formula.

[0019] The control module is further configured to control the first mechanical arm to place the reagent bottle on the weighing scale into the heating and mixing device after the powder injection device and the liquid injection device complete the operation.

[0020] The heating and mixing device is configured to heat and mix the powder and the solvent in the reagent bottle.

[0021] The control module is further configured to control the first mechanical arm to place the reagent bottle after heating and mixing on the visual recognition device.

[0022] The visual recognition device is configured to identify whether the liquid in the reagent bottle is qualified.

[0023] The control module is further configured to control the first mechanical arm to take out the qualified reagent bottle to obtain the configured electrolyte.

[0024] In an embodiment, the test parameters include physical property parameters representing physical properties of the electrolyte and electrochemical parameters representing electrochemical properties of the electrolyte.

[0025] The electrolyte test module includes a second mechanical arm, a plurality of sub-liquid containers, an electrochemical device assembly device, a physical property test device, and an electrochemical test device.

[0026] The control module is configured to control the second mechanical arm to sub-pack the configured electrolyte into the plurality of sub-liquid containers.

[0027] The control module is further configured to control the second mechanical arm to send part of the sub-liquid containers to the physical property test device for physical property test to obtain the physical property parameters of the electrolyte.

[0028] The control module is further configured to control the second mechanical arm to send another part of the sub-liquid containers to the electrochemical device assembly device for assembly to obtain electrochemical devices, and control the electrochemical test device to perform electrochemical test on the electrochemical devices to obtain the electrochemical parameters of the electrolyte.

[0029] In addition, to achieve the above object, the present application further provides a storage medium, which is a computer readable storage medium, and a computer program is stored on the storage medium. The computer program is executed by a processor to implement the steps of the electrolyte recommendation method described above.

[0030] In addition, to achieve the above object, the present application further provides a computer program product, which includes a computer program. The computer program is executed by a processor to implement the steps of the electrolyte recommendation method described above.

[0031] The one or more technical solutions provided in the present application have at least the following technical effects:

[0032] In the embodiment, based on a plurality of electrolyte formulations, a plurality of test parameters of electrolytes, and theoretical parameters of a plurality of electrolyte formulations, an electrolyte formulation meeting preset requirements is recommended by a pre-trained machine learning model. The pre-trained machine learning model learns the correlation between electrolyte formulations and test parameters and theoretical parameters, and recommends an electrolyte formulation meeting preset requirements according to the correlation. In this way, a test personnel does not need to test a large number of different formulations to develop an electrolyte formulation, which is efficient and avoids the problems of poor objectivity and inaccuracy caused by completely relying on subjective understanding and experience analysis, and provides a new method that can improve the development efficiency and accuracy of electrolyte formulations. BRIEF DESCRIPTION OF DRAWINGS

[0033] The accompanying drawings, which are incorporated herein and constitute part of the specification, illustrate embodiments consistent with the application and, together with the description, serve to explain the principles of the application.

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings can also provide the basis for obtaining other drawings for those skilled in the art without any creative work.

[0035] Figure 1 A flowchart provided for the electrolyte recommendation method embodiment one of the present application;

[0036] Figure 2 A flowchart provided for the electrolyte recommendation method embodiment two of the present application;

[0037] Figure 3 A structural diagram provided for the electrolyte recommendation system embodiment one of the present application;

[0038] Figure 4 A structural diagram provided for the electrolyte recommendation system embodiment two of the present application.

[0039] The purpose of the present application, functional characteristics and advantages will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0040] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application, and are not used to limit the present application.

[0041] In order to better understand the technical solutions of the present application, the following will be described in detail in combination with the drawings of the specification and the specific embodiments.

[0042] The embodiments of the present application aim to provide a new method which can improve the efficiency and accuracy of electrolyte formula development.

[0043] With the wide use of batteries, the research and preparation of high-performance batteries have attracted more and more attention. Electrolyte is one of the main components of the battery, which plays a role in transferring the charge between the positive and negative electrodes. The performance of the electrolyte itself and the interface condition formed by the electrolyte and the positive and negative electrodes of the battery have a great influence on the performance of the battery. An electrolyte with excellent performance requires good chemical stability, electrochemical window and electrical conductivity parameters. In order to obtain better battery performance, the electrolyte in the battery needs to be adjusted to obtain a higher matching degree of component ratio with the positive and negative electrodes.

[0044] Currently, the development of electrolyte formula mainly relies on manual experiments, which needs to configure and test a large number of different formulas, and the material test and characterization cycle is long, resulting in low efficiency; and the experimental data mainly depends on the analysis of the experimenters, which is highly subjective, and the accuracy of the developed electrolyte formula is not high. Therefore, there is an urgent need for a new method that can improve the development efficiency and accuracy of the electrolyte formula.

[0045] The present application provides an electrolyte recommendation method, which is based on a plurality of electrolyte formulas, a plurality of test parameters of electrolytes, and a plurality of theoretical parameters of electrolyte formulas, and recommends an electrolyte formula meeting a preset requirement through a pre-trained machine learning model. The pre-trained machine learning model learns the correlation between the electrolyte formula and the test parameters and / or the theoretical parameters, and recommends an electrolyte formula meeting the preset requirement according to the correlation. In this way, the tester does not need to test a large number of different formulas to develop the electrolyte formula, which is highly efficient, and avoids the problems of poor objectivity and inaccuracy caused by completely relying on manual subjective understanding and experience analysis, thereby providing a new method that can improve the development efficiency and accuracy of the electrolyte formula.

[0046] Based on this, the present application provides an electrolyte recommendation method, which is based on a plurality of electrolyte formulas, a plurality of test parameters of electrolytes, and a plurality of theoretical parameters of electrolyte formulas, and recommends an electrolyte formula meeting a preset requirement through a pre-trained machine learning model. The pre-trained machine learning model learns the correlation between the electrolyte formula and the test parameters and / or the theoretical parameters, and recommends an electrolyte formula meeting the preset requirement according to the correlation. In this way, the tester does not need to test a large number of different formulas to develop the electrolyte formula, which is highly efficient, and avoids the problems of poor objectivity and inaccuracy caused by completely relying on manual subjective understanding and experience analysis, thereby providing a new method that can improve the development efficiency and accuracy of the electrolyte formula. Figure 1 Figure 1 FIG. 1 is a flowchart of the first embodiment of the electrolyte recommendation method of the present application.

[0047] In this embodiment, the electrolyte recommendation method includes steps S10-S30:

[0048] Step S10: Configure a plurality of electrolytes based on a plurality of electrolyte formulas.

[0049] In this embodiment, the electrolyte formula includes a plurality of chemical components and the proportions of the plurality of chemical components for preparing the electrolyte. The plurality of chemical components generally include a solvent and a solute, wherein the solute is generally a powder, such as various types of lithium salts, and the solvent is generally a liquid, which can be composed of one or more solvents, such as ethylene carbonate (EC), diethyl carbonate (DEC), dimethyl carbonate (DMC), and methyl ethyl carbonate (EMC). The proportion of the plurality of chemical components is the proportion of the solvent and the solute, and if the solvent is multiple, the proportion of the multiple solvents also needs to be determined. The designer pre-designs a plurality of electrolyte formulas, and then configures a plurality of electrolytes based on the pre-designed plurality of electrolyte formulas.

[0050] Step S20: Obtain the test parameters of each electrolyte and the theoretical parameters of each electrolyte formula, respectively.

[0051] ​The test parameters in this embodiment are parameters representing the performance of the electrolyte obtained by testing the electrolyte, and the theoretical parameters are parameters representing the performance of the electrolyte obtained by theoretically calculating the electrolyte formula. In this embodiment, the test parameters representing the performance of the electrolyte and the theoretical parameters representing the performance of the electrolyte are obtained for each electrolyte, so as to avoid the inaccuracy of determining the electrolyte formula only according to the test parameters or only according to the theoretical parameters.

[0052] In step S30, based on the plurality of electrolyte formulas, the test parameters of the plurality of electrolytes, and the theoretical parameters of the plurality of electrolyte formulas, an electrolyte formula meeting the preset requirement is recommended by the pre-trained machine learning model.

[0053] The machine learning model includes an association relationship, which represents the association relationship between the electrolyte formula and the test parameters and / or the theoretical parameters. The preset requirement includes that the test parameters and / or the theoretical parameters are within a preset range. Specifically, only the test parameters are within the preset range, or only the theoretical parameters are within the preset range, or both the test parameters and the theoretical parameters are within the preset range. The specific range value in the preset requirement can be set as needed, which is not limited in this embodiment.

[0054] For example, if the preset requirement only includes that the test parameters are within the preset range.

[0055] If the test parameters include multiple parameters, a range threshold is set for each test parameter. For example, if the test parameters include coulomb efficiency and conductivity of the electrolyte, a range threshold is set for each test parameter, for example, a first preset range is set for the coulomb efficiency, and a second preset range is set for the conductivity of the electrolyte. When the coulomb efficiency is within the first preset range and the conductivity of the electrolyte is within the second preset range, it is considered that the test parameters meet the preset requirement.

[0056] For example, since high coulomb efficiency means that the battery loses less during charging and discharging, the first preset range is generally set to be large, for example, the first preset range can be 99% to 99.5%. Since the reduction of the conductivity of the electrolyte will directly affect the internal resistance of the battery, and then affect the efficiency and performance of the battery, the second preset range should not be too low, for example, the second preset range can be between 10 mS / cm and 50 mS / cm, where mS / cm is millisiemens per centimeter.

[0057] For example, if the preset requirement only includes that the theoretical parameters are within the preset range.

[0058] If the theoretical parameters include multiple parameters, a range threshold is set for each theoretical parameter. For example, if the theoretical parameters include dielectric constant and ion diffusion coefficient, a range threshold is set for each theoretical parameter, for example, a third preset range is set for the dielectric constant, and a fourth preset range is set for the ion diffusion coefficient. When the dielectric constant is within the third preset range and the ion diffusion coefficient is within the fourth preset range, it is considered that the theoretical parameters meet the preset requirements.

[0059] For example, because electrolyte with high dielectric constant can better isolate charges and reduce the interaction between ions, the third preset range is generally set to be large, for example, the third preset range can be 6-6.5. Because high ion diffusion coefficient is usually related to better battery performance, especially in application scenarios that require high power output and fast charging and discharging, the fourth preset range should not be too low, for example, the fourth preset range can be 5x 10^-11 m2 / s-1x 10^-10 m2 / s, where m2 / s is square meter per second.

[0060] The machine learning model pre-trained in the embodiment can learn the correlation between the electrolyte formula and the test parameters and / or the theoretical parameters, and recommend an electrolyte formula that meets the preset requirements according to the correlation. In this way, the test personnel do not need to test a large number of different formulas to develop an electrolyte formula, which is more efficient, and avoids the problems of poor objectivity and inaccuracy caused by completely relying on subjective understanding and experience analysis, and provides a new method that can improve the efficiency and accuracy of electrolyte formula development.

[0061] The multiple electrolyte formulas in the embodiment of the application include multiple electrolyte formulas composed of different types of chemical components in different proportions.

[0062] The electrolyte formulas are described below with specific examples.

[0063] Formula 1: LiFSI:FEC:EMC (1:6:4), Formula 2: LiBF4:FEC:EMC (1:6:4), Formula 3: LiFSI:FEC:EMC (1:7:3), Formula 4: LiBF4:FEC:EMC (1:7:3).

[0064] In the above four electrolyte formulations, in the electrolyte with fluoro ethylene carbonate (FEC) and ethyl methyl carbonat (EMC) as mixed solvents, the designer gives four different electrolyte formulations, the first one takes lithium bis(trifluoromethanesulfonyl)imide (LiFSI) as lithium salt, wherein the ratio of LiFSI:FEC:EMC is 1:6:4; the second one takes lithium tetrafluoroborate (LiBF4) as lithium salt, wherein the ratio of LiBF4:FEC:EMC is also 1:6:4; the third one also takes LiFSI as lithium salt, the ratio of LiFSI:FEC:EMC is 1:7:3; the fourth one also takes LiBF4 as lithium salt, the ratio of LiBF4:FEC:EMC is 1:7:3.

[0065] In an implementable embodiment, the test parameters include physical property parameters characterizing physical properties of the electrolytes and electrochemical parameters characterizing electrochemical properties of the electrolytes.

[0066] As shown in FIG. 2, step S20 can include steps S21-S22: Figure 2

[0067] Step S21, physical property testing and electrochemical testing are performed on each electrolyte to obtain physical property parameters and electrochemical parameters of each electrolyte.

[0068] Specifically, the physical property parameters mainly describe the physical properties of the electrolyte without involving electrochemical reactions. These parameters can usually be measured by non-electrochemical methods, for example, by physical property testing.

[0069] The physical property testing includes one or more of viscosity testing of the electrolyte at different temperatures, functional group testing of the electrolyte, solvent chemical structure testing of the electrolyte, saturation pressure value testing of the electrolyte at different temperatures, and conductivity testing of the electrolyte.

[0070] Among them, the physical property parameter obtained by viscosity testing of the electrolyte at different temperatures is that the viscosity value of the electrolyte at different temperatures can be measured. Viscosity testing is a parameter used to measure the flow resistance of the electrolyte. Viscosity testing at different temperatures can understand how the flow performance of the electrolyte changes with temperature. Viscosity is usually measured using a viscometer (such as a rotary viscometer or a capillary viscometer). Viscosity is greatly affected by temperature, and generally the viscosity of a liquid decreases as the temperature increases.

[0071] ​The functional group test of the electrolyte results in the physical property parameter: specific functional groups of the electrolyte. Functional group testing is a test that determines the type and content of functional groups present in the electrolyte through chemical or instrumental analysis methods. Common methods include infrared spectroscopy, nuclear magnetic resonance spectroscopy, mass spectrometry, etc. These tests help understand the chemical structure of the electrolyte and its possible reaction characteristics.

[0072] The solvent chemical structure test of the electrolyte results in the physical property parameter: the molecular structure and composition of the solvent in the electrolyte. Solvent chemical structure testing is used to analyze the molecular structure and composition of the solvent in the electrolyte. This usually involves the use of techniques such as nuclear magnetic resonance spectroscopy, mass spectrometry, infrared spectroscopy, etc. to identify the types of solvents and their chemical environment.

[0073] The saturated gas pressure value test of the electrolyte at different temperatures results in the physical property parameter: the volatility of the electrolyte at different temperatures. The saturated gas pressure value test measures the pressure of the electrolyte when it reaches dynamic equilibrium at a certain temperature. This test helps understand the volatility of the electrolyte, which is very important for the safety and storage conditions of the electrolyte. A saturated vapor pressure instrument is usually used for testing.

[0074] The conductivity test of the electrolyte results in the physical property parameter: the conductivity value of the electrolyte. Conductivity testing is a parameter that measures the electrical conductivity of the electrolyte, which reflects the concentration and mobility of ions in the electrolyte. Conductivity is usually measured by a conductivity meter and can be performed at different frequencies and temperatures. Conductivity is a key indicator for evaluating the performance of electrolytes, especially in battery and electrochemical applications.

[0075] These tests are indispensable steps in the development and application of electrolytes, providing important information about the behavior and performance of electrolytes under different conditions, helping to optimize electrolyte formulations and improve their use conditions.

[0076] It is worth noting that when performing physical property tests of the electrolyte, in order to ensure the accuracy of the test results, the same electrolyte can only be tested once, and the same electrolyte cannot be used for multiple tests.

[0077] Electrochemical parameters describe the behavior and performance of the electrolyte in electrochemical reactions, which are usually measured by electrochemical test methods. Electrochemical tests include one or more of the following: coulombic efficiency test, limiting current density test, electrochemical window value test, ion diffusion coefficient test, and AC impedance test.

[0078] Among them, the electrochemical parameter obtained by coulomb efficiency test is: coulomb efficiency. Coulomb efficiency refers to the ratio of the actual charge stored and released during the charging and discharging process of the battery to the theoretical charge, usually expressed in percentage. Coulomb efficiency test is carried out by measuring the actual capacity of the battery in continuous charging and discharging cycles. High coulomb efficiency means that the battery loses less during charging and discharging, which is an important parameter for evaluating battery performance and life.

[0079] The electrochemical parameter obtained by limiting current density test is: limiting current density. Limiting current density refers to the maximum current density that an electrode can withstand without serious side reactions or damage during electrochemical processes. This test involves gradually increasing the current density until a significant decrease in electrode performance or instability is observed. Limiting current density test helps determine the optimal operating range of the battery or electrochemical device.

[0080] The electrochemical parameter obtained by electrochemical window value test is: electrochemical window value. Electrochemical window refers to the voltage range within which the electrolyte does not undergo decomposition reactions during electrochemical processes. Electrochemical window value test is usually measured by linear sweep voltammetry or cyclic voltammetry, that is, scanning the electrode potential within a certain voltage range and observing the current response of the electrolyte to determine the stable voltage range of the electrolyte.

[0081] The electrochemical parameter obtained by ion diffusion coefficient test is: ion diffusion coefficient. Ion diffusion coefficient is a parameter describing the diffusion rate of ions in the electrolyte, usually tested by electrochemical impedance spectroscopy or chronopotentiometry. These methods can help researchers understand the migration speed of ions in the electrolyte, which is crucial for the design and application of batteries.

[0082] The electrochemical parameter obtained by alternating current impedance test is: alternating current impedance value. Alternating current impedance test is an electrochemical test technique used to measure the impedance of electrode materials and electrolytes under alternating current signals. By measuring impedance and phase angle at different frequencies, information about electrode reaction kinetics, charge transfer resistance, double-layer capacitance, and ion diffusion of electrochemical processes can be obtained. Impedance spectroscopy is a commonly used method that analyzes Nyquist or Bode plots to obtain this information.

[0083] In this embodiment, the electrolyte needs to be assembled into an electrochemical device before electrochemical testing, such as a simple battery. For example, the electrolyte is loaded into a simple device with a positive electrode and a negative electrode to assemble a simple battery.

[0084] In order to ensure the accuracy of the test results, the same electrochemical device can only be tested once when performing electrochemical tests of electrolyte. Therefore, it is necessary to assemble the configured electrolyte into multiple identical electrochemical devices to perform different tests on each electrochemical device.

[0085] It is worth noting that the above-mentioned electrochemical tests and physical property tests are only illustrative examples, and other electrochemical tests or physical property tests are also included in the embodiments of the present application.

[0086] Step S22, based on first principles and / or molecular dynamics simulation, theoretical calculation is carried out on each electrolyte formula, and the theoretical parameters of each electrolyte formula are obtained.

[0087] First-principles calculation, also known as ab initio calculation, is based on the principles of quantum mechanics, without any experimental data or empirical parameters to describe the electronic structure of materials. This method calculates the wave function of electrons by solving the Schrödinger equation, thereby obtaining the electronic structure, energy, bond length, bond angle, and other properties of materials.

[0088] The process of calculating the theoretical parameters of electrolyte formula based on first principles can be shown as follows: (1) Structure optimization: First, the geometric structure of the molecules in the electrolyte (such as solvent, solute) is optimized to find their stable configuration. (2) Electronic structure analysis: Calculate the electronic structure of the molecule, including band structure, density of states, charge distribution, etc., to understand the electronic interaction within the molecule. (3) Reaction path analysis: By calculating the transition state and reaction energy barrier of the reaction, the chemical reaction path and reaction activity in the electrolyte can be predicted. (4) Property prediction: Based on the electronic structure calculation, the electrochemical window, ion migration energy, ion solvation energy, and other properties of the electrolyte can be predicted.

[0089] Molecular dynamics simulation is a computer simulation technique based on Newton's law of motion, which simulates the motion of atoms and molecules by solving the motion equations of all atoms in the system. In the simulation process, the interaction force between atoms is usually described by an empirical potential function or a force field calculated by first principles. Molecular dynamics can be used to study the dynamic properties of materials, such as diffusion coefficient, viscosity, thermal conductivity, etc.

[0090] The process of calculating theoretical parameters of electrolyte formulations based on molecular dynamics simulation can be as follows: (1) Model construction: Construct a molecular model of the electrolyte, including the solvent, solute, and possible ion pairs. (2) Force field parameterization: Choose appropriate force field parameters or use first-principle calculations to obtain parameters to describe atomic interactions. (3) Simulation running: Run molecular dynamics simulation at a specific temperature and pressure, and observe the behavior of the system. (4) Property calculation: Extract physical properties of the electrolyte from the simulation trajectory, such as density, viscosity, ion diffusion coefficient, etc.

[0091] By combining first-principle and molecular dynamics simulation, the behavior of electrolytes can be understood at the atomic and molecular level, providing theoretical guidance for the design and optimization of electrolyte formulations. These calculation methods can greatly reduce the experimental workload and accelerate the research and development process of new materials.

[0092] In one possible implementation, the theoretical parameters include one or more of the following: density, ion binding energy, molecular dissociation energy, molecular surface electrostatic potential, dielectric constant, solvent donor number, solvent acceptor number, ion diffusion coefficient, and space charge density.

[0093] Density: refers to the mass per unit volume, usually expressed in kilograms per cubic meter (kg / m3) or grams per cubic centimeter (g / cm3). For electrolytes, density is an important physical parameter that measures the mass-to-volume ratio.

[0094] Ion binding energy: refers to the energy of interaction between ions when forming ion pairs. It reflects the stability and interaction strength of ions in the electrolyte.

[0095] Molecular dissociation energy: refers to the energy required to decompose a molecule into atoms or smaller molecules. In electrolytes, molecular dissociation energy can affect the degree of dissociation of electrolytes and conductivity.

[0096] Molecular surface electrostatic potential: refers to the potential difference on the surface of a molecule due to uneven charge distribution. It can affect the interaction between molecules and the physical and chemical properties of the electrolyte.

[0097] Dielectric constant: the dielectric constant is a measure of the polarization ability of a material in an electric field, reflecting the degree of response of the material to the electric field. Electrolytes with high dielectric constant can better isolate charges and reduce ion-ion interactions.

[0098] Solvent donor number (Donor Number, DN): is a parameter that measures the ability of a solvent to donate electron pairs, commonly used to describe the basicity of a solvent. It can be determined by specific experimental methods.

[0099] Acceptor Number (AN): A parameter that measures the ability of a solvent to accept electron pairs, often used to describe the acidity of a solvent. It can also be determined through specific experimental methods.

[0100] Ionic Diffusion Coefficient: A physical quantity that describes the rate of diffusion of ions in an electrolyte, usually expressed in square centimeters per second (cm2 / s). It is an important parameter for measuring the ion transport performance of an electrolyte.

[0101] Space Charge Density: Refers to the accumulation of electric charge in an electrolyte due to different ion migration speeds or electrolyte concentration gradients, which affects the conductivity of the electrolyte and the electrochemical behavior near the electrode.

[0102] These theoretical parameters are important indicators in the performance evaluation of electrolytes, which can provide information about the physical, chemical and electrochemical properties of electrolytes, and are crucial for the design and application of electrolytes. It is worth mentioning that the above theoretical parameters are only for example, other theoretical parameters are also included in the embodiments of the present application.

[0103] It is worth mentioning that some parameters of the electrolyte can be obtained by testing or by theoretical calculation. For such parameters, theoretical calculation can be preferred to avoid the time cost of testing.

[0104] In a feasible implementation, the pre-trained machine learning model is obtained by the following method.

[0105] (1) Collect sample data, including multiple sample electrolyte formulations, sample test parameters and sample theoretical parameters corresponding to multiple sample electrolyte formulations.

[0106] Suppose a new type of lithium ion battery electrolyte is being developed, the following data has been collected:

[0107] Sample electrolyte formulation: 100 different electrolyte formulations, each including different proportions of solvents, solutes and additives.

[0108] Sample test parameters: Actual test results corresponding to each electrolyte formulation, such as density, ion diffusion coefficient, conductivity, etc.

[0109] Sample theoretical parameters: Theoretical parameters obtained through first-principle calculations and molecular dynamics simulations, such as molecular surface electrostatic potential, ion binding energy, etc.

[0110] (2) Use a portion of the sample data to train the machine learning model to obtain the correlation.

[0111] From the sample electrolyte formula, 80 kinds of electrolyte formula and their corresponding test parameters and theoretical parameters are selected as the training data set. The training data set is used to train the model, and the correlation between the electrolyte formula and the performance parameters is established. The correlation between the electrolyte formula and the test parameters and / or theoretical parameters is represented by the correlation between the electrolyte formula and the test parameters and / or theoretical parameters. The correlation includes at least the positive correlation between the electrolyte formula and the test parameters and / or theoretical parameters, and the negative correlation between the electrolyte formula and the test parameters and / or theoretical parameters. For example, the molar mass of a chemical substance in the electrolyte formula increases, and the test parameter becomes larger or smaller, and the theoretical parameter becomes larger or smaller.

[0112] (3) Use another part of the sample data that does not participate in training to evaluate the performance of the machine learning model. According to the evaluation result of the performance of the machine learning model, the machine learning model is optimized until the performance of the machine learning model meets the performance threshold.

[0113] From the sample electrolyte formula, 20 kinds of electrolyte formula and their parameters that do not participate in training are reserved as test data set to evaluate the performance of the model. The performance of the model can be evaluated using the following indicators: accuracy of the model, number of training times of the model, mean square error, determination coefficient, mean absolute error, etc. For example, the accuracy of the machine learning model is not less than 95% to meet the performance threshold.

[0114] If it is found that the performance of the model has not reached the expected performance threshold, the following are some possible optimization measures: A, feature selection: by analyzing the feature importance, remove or add some features to improve the performance of the model. B, adjust the hyperparameters: adjust the hyperparameters of the model, such as the depth of the tree, the learning rate, etc. through grid search or random search. C, use more data: if possible, collect more data to increase the generalization ability of the model. D, model fusion: try to fuse the prediction results of multiple models to improve the prediction accuracy.

[0115] After several iterations of optimization, we finally get a machine learning model whose performance meets the threshold. This model can now be used to predict the performance parameters of new electrolyte formulas without experimental testing, thereby speeding up the development process of electrolytes.

[0116] Example: Suppose we have trained a model to predict the conductivity of electrolyte. After optimization, the performance of our model on the test set (such as accuracy) is not less than 95%, indicating that the model has good prediction ability. Now, a new electrolyte formula is proposed, and we use this model to predict its conductivity, with the result showing 5 mS / cm. This prediction result can be used as a reference for recommending electrolyte formulas.

[0117] The following will illustrate the process of how the pre-trained machine learning model recommends electrolyte formulations by taking specific electrolyte formulations as examples.

[0118] Firstly, a plurality of electrolyte formulations and the parameters obtained are shown in Table 1.

[0119] Table 1

[0120]

[0121]

[0122] For example, in a sample electrolyte with fluoro ethylene carbonate (FEC) and ethyl methyl carbonat (EMC) as mixed solvents, the designer gives four different electrolyte formulations. The first one uses lithium bis(trifluoromethanesulfonyl)imide (LiFSI) as lithium salt, and the ratio of LiFSI:FEC:EMC is 1:6:4. The second one uses lithium tetrafluoroborate (LiBF4) as lithium salt, and the ratio of LiBF4:FEC:EMC is also 1:6:4. The third one also uses LiFSI as lithium salt, and the ratio of LiFSI:FEC:EMC is 1:7:3. The fourth one also uses LiBF4 as lithium salt, and the ratio of LiBF4:FEC:EMC is 1:7:3.

[0123] The third and fourth electrolyte formulations are compared to the fixed molar ratio of mixed solvents (FEC:EMC, 7:3) and different types of lithium salts (such as LiFSI and LiBF4) are added. The first and third electrolyte formulations are compared to the fixed molar ratio of lithium salt (LiFSI), and the ratio of each solvent in the mixed solvent is changed, such as FEC:EMC in the third formulation is 7:3, and FEC:EMC in the first formulation is 6:4. The second and fourth electrolyte formulations are compared to the fixed molar ratio of lithium salt (LiBF4), and the ratio of each solvent in the mixed solvent is changed, such as FEC:EMC in the fourth formulation is 7:3, and FEC:EMC in the second formulation is 6:4. As can be seen from the above Table 1, the test parameters and theoretical parameters obtained from the four electrolyte formulations are significantly different.

[0124] Secondly, the pre-trained machine learning model learns the correlation between electrolyte formulations and various parameters, including at least positive correlation and negative correlation.

[0125] As shown in Table 1 above, in the electrolyte system of FEC and EMC binary solvent, under the same molar concentration of lithium salt substance, the electrolyte containing LiFSI has higher coulomb efficiency, higher dielectric constant, lower viscosity and lower saturated vapor pressure value than the electrolyte containing LiBF4. That is, under the same molar concentration of lithium salt substance, LiFSI has a positive correlation with coulomb efficiency and dielectric constant, and a negative correlation with viscosity and saturated vapor pressure value; LiBF4 has a negative correlation with coulomb efficiency and coulomb efficiency, and a positive correlation with viscosity and saturated vapor pressure value.

[0126] When the type and molar concentration of lithium salt substance are determined (for example, the lithium salt is LiFSI and the molar concentration is 1; or, the lithium salt is LiBF4 and the molar concentration is 1), increasing the molar ratio of the solvent formula FEC / EMC can increase the viscosity of the electrolyte, improve the coulomb efficiency, and reduce the saturated vapor pressure and dielectric constant. That is, the molar ratio of the solvent formula FEC / EMC has a positive correlation with the viscosity and coulomb efficiency of the electrolyte, and a negative correlation with the saturated vapor pressure and dielectric constant.

[0127] Finally, after learning the correlation between the electrolyte formula and each parameter, and obtaining the preset requirements of the test parameters and / or the theoretical parameters within the preset range, the reasonable proportion of the solvent and solute in the electrolyte formula can be deduced, thereby obtaining the electrolyte formula that meets the preset requirements.

[0128] For example: after determining the solvent and solute of the electrolyte, the electrolyte formula closest to the preset requirements is first found in the existing electrolyte formula. For example: if the solvent of the electrolyte is determined to be a mixed solvent of FEC and EMC, the solute is LiFSI, and the preset requirement is that the coulomb efficiency of the electrolyte is not less than 99% and the viscosity is not higher than 1, at this time, according to Table 1, the electrolyte formula closest to the preset requirements is determined to be the first LiFSI:FEC:EMC (1:6:4).

[0129] Based on the above-learned correlation, when the type and molar concentration of lithium salt substance are determined, the molar ratio of the solvent formula FEC / EMC has a positive correlation with the viscosity of the electrolyte, therefore, through the learned correlation, it is known that the molar ratio of FEC / EMC needs to be reduced to reduce the viscosity of the electrolyte, in this way, through the correlation, the electrolyte formula closest to the preset requirements can be adjusted to obtain a suitable molar ratio of FEC / EMC, thereby obtaining an electrolyte formula that meets the preset requirements.

[0130] The following will illustrate how to solve the problem of how to select lithium salt and how to determine the molar ratio of the solvent when the mixed solvent is FEC and EMC by machine learning, by combining the above-mentioned design of four kinds of electrolytes.

[0131] (1) Through analyzing the test parameters of the four electrolytes and the theoretical parameters of the four electrolyte formulations, the following rules can be found: in the mixed solvent of FEC and EMC, adding LiFSI has lower viscosity (the lower the viscosity, the more conducive to reducing the ion transmission resistance and improving the rate performance of the battery), lower saturated vapor pressure value (the electrolyte with low saturated vapor pressure value has low volatility, and its high temperature safety performance is better), higher dielectric constant (the larger the dielectric constant, the smaller the force between lithium ions and anions, the easier the dissociation of lithium salt, the more free ions, the higher the conductivity, and the better the rate performance), and higher coulomb efficiency. Therefore, it can be learned that LiFSI is preferred as the electrolyte lithium salt in the mixed solvent of FEC and EMC;

[0132] (2) Through analyzing the test parameters of the four electrolytes and the theoretical parameters of the four electrolyte formulations, the following rules can be found: when LiFSI is selected as the lithium salt, increasing the molar ratio of FEC / EMC is conducive to improving the high temperature safety of the electrolyte (reducing the saturated vapor pressure) and the coulomb efficiency; however, increasing the molar ratio of FEC / EMC will increase the viscosity of the electrolyte and reduce the dielectric constant, which is not conducive to the improvement of the rate performance. Therefore, the molar ratio of the solvents in the electrolyte can be guided in reverse according to the actual use scene and function of the electrolyte.

[0133] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0134] The present application also provides an electrolyte recommendation system, as shown in Figure 3 The system includes a control module 10, an electrolyte configuration module 20 and an electrolyte test module 30 connected with the control module 10, and a computing device 40 connected with the electrolyte test module 30.

[0135] The control module 10 is used to control the electrolyte configuration module 20 to configure a plurality of electrolytes based on a plurality of electrolyte formulations, and the electrolyte formulation includes a plurality of chemical components for preparing the electrolyte and a ratio of the plurality of chemical components.

[0136] The control module 10 is also used to control the electrolyte test module 30 to test each electrolyte configured, to obtain test parameters of each electrolyte, and the test parameters are parameters obtained by testing the electrolyte, which represent the performance of the electrolyte.

[0137] The computing device 40 is configured to obtain the test parameters of each electrolyte, and is further configured to perform theoretical calculation based on each electrolyte formula to obtain the theoretical parameters of each electrolyte formula, the theoretical parameters being parameters obtained by performing theoretical calculation on the electrolyte formula and representing the performance of the electrolyte.

[0138] The computing device 40 is further configured to recommend an electrolyte formula meeting preset requirements by using a pre-trained machine learning model based on the plurality of electrolyte formulas, the test parameters of the plurality of electrolytes, and the theoretical parameters of the plurality of electrolyte formulas, wherein the machine learning model comprises a correlation relationship representing the correlation relationship between the electrolyte formula and the test parameters and / or the theoretical parameters, and the preset requirements comprise that the test parameters and / or the theoretical parameters are within a preset range.

[0139] Specifically, the control module 10 starts a work flow after obtaining the electrolyte formula and the test method input by the designer.

[0140] Firstly, the control module 10 controls the electrolyte configuration module 20 to configure a plurality of electrolytes based on the plurality of electrolyte formulas; secondly, the control module 10 controls the electrolyte test module 30 to test each electrolyte configured to obtain the test parameters of each electrolyte.

[0141] Then, the computing device 40 obtains the test parameters of each electrolyte. The computing device 40 further performs theoretical calculation based on each electrolyte formula to obtain the theoretical parameters of each electrolyte formula.

[0142] Finally, the computing device 40 recommends an electrolyte formula meeting preset requirements by using a pre-trained machine learning model based on the test parameters of the plurality of electrolytes and the theoretical parameters of the plurality of electrolyte formulas.

[0143] The computing device 40 constructs an electrolyte formula-test parameter-theoretical parameter database corresponding to each formula based on the test parameters of the plurality of electrolytes and the theoretical parameters of the plurality of electrolyte formulas. The correlation relationship between the electrolyte formula and the test parameters and / or the theoretical parameters is obtained by machine learning of the multi-dimensional data in the database through AI intelligent calculation. The machine learning method comprises Naive Bayes, random forest, decision tree, neural network, etc.

[0144] The computing device 40 intelligently recommends the electrolyte formula designed by the designer according to the correlation relationship between the electrolyte formula-test parameter-theoretical parameter, and adjusts the existing electrolyte formula to obtain an electrolyte formula meeting preset requirements. The computing device 40 feeds back the electrolyte formula meeting the preset requirements to the electrolyte configuration module 20 to guide the formula design of the electrolyte configuration module 20.

[0145] The computing device 40 recommends an electrolyte formula meeting preset requirements based on a plurality of electrolyte formulas, test parameters of a plurality of electrolyte formulas, and theoretical parameters of a plurality of electrolyte formulas through a pre-trained machine learning model, learns the correlation between the electrolyte formula and the test parameters and / or the theoretical parameters by the pre-trained machine learning model, and recommends the electrolyte formula meeting the preset requirements according to the correlation, so that the test personnel do not need to test a large number of different formulas to develop the electrolyte formula, the efficiency is higher, and the problems of poor objectivity and inaccuracy caused by completely relying on subjective understanding and experience analysis are avoided, thereby providing a new method that can improve the electrolyte formula development efficiency and development accuracy.

[0146] In a feasible implementation, referring to Figure 4 , the electrolyte configuration module 20 includes a first mechanical arm 21, a reagent bottle 22, a powder injection device 23, a liquid injection device 24, a weighing scale 25, a heating and mixing device 26, and a visual recognition device 27.

[0147] The control module 10 is configured to control the first mechanical arm 21 to take the reagent bottle 22 and place the reagent bottle 22 on the weighing scale 25, and the powder injection device 23 is configured to inject a powder into the reagent bottle 22 on the weighing scale 25 based on an electrolyte formula of a plurality of electrolyte formulas, and the liquid injection device 24 is configured to inject a solvent into the reagent bottle 22 on the weighing scale 25 based on the electrolyte formula.

[0148] The control module 10 is further configured to control the first mechanical arm 21 to place the reagent bottle 22 on the weighing scale 25 into the heating and mixing device 26 after the powder injection device 23 and the liquid injection device 24 complete the operation, and the heating and mixing device 26 is configured to heat and mix the powder and the solvent in the reagent bottle 22.

[0149] The control module 10 is further configured to control the first mechanical arm 21 to place the reagent bottle 22 after heating and mixing in the visual recognition device 27, and the visual recognition device 27 is configured to identify whether the electrolyte in the reagent bottle 22 is qualified.

[0150] The control module 10 is further configured to control the first mechanical arm 21 to take out the qualified reagent bottle 22 to obtain the configured electrolyte.

[0151] Specifically, as shown in Figure 4 , the control module 10 controls the working process of the electrolyte configuration module 20 to configure the electrolyte as follows:

[0152] (1) Based on the electrolyte formula designed by the artificial, the artificial prepares the powder material, the solvent, and the reagent bottle 22 required in the electrolyte formula, and transports the above-mentioned materials to the powder storage device area, the liquid storage device area, and the reagent bottle 22 storage area through the first mechanical arm 21.

[0153] (2) Input the artificial designed electrolyte formula and test method in the control module 10, and start the system workflow.

[0154] (3) The first mechanical arm 21 grabs the empty reagent bottle 22 in the reagent bottle 22 storage area and transfers it to the weighing scale 25.

[0155] (4) The powder injection device 23 includes a plurality of powder storage modules 231, and a grinding module 232 and an automatic powder injection module 233 corresponding to each powder storage module 231. Different powders are automatically transferred from different powder storage modules to the corresponding grinding module 232 for grinding, and then to the corresponding automatic powder injection module 233. The automatic powder injection module 233 automatically injects powder into the empty reagent bottle 22 on the weighing scale 25, and stops injection when the weighing reaches the set target value.

[0156] (5) The liquid injection device 24 includes a plurality of liquid storage modules 241, and an automatic liquid injection module 242 corresponding to each liquid storage module 241. After the powder injection is completed, different solvents are automatically transferred from different liquid storage modules 241 to the corresponding automatic liquid injection module 242, and the automatic liquid injection module 242 automatically injects liquid into the reagent bottle 22 on the weighing scale 25. The injection stops when the weighing reaches the set target value.

[0157] (6) The first mechanical arm 21 picks up the reagent bottle 22 containing powder and solvent and transfers it to the heating and mixing device 26. The heating and mixing device 26 stirs and mixes according to the set time, temperature, and frequency.

[0158] (7) The stirred reagent bottle 22 is transferred by the first mechanical arm 21 to the visual recognition device 27 for sedimentation and layering detection. The heating and mixing device 26 stirs and mixes according to the set time, temperature, and frequency. After the set stirring time is reached, visual judgment is performed. If the visual judgment passes, the electrolyte is determined to be qualified and can proceed to the next test environment. If the visual judgment fails, the electrolyte is determined to be unqualified. After all reagent bottles 22 are tested on the same day, they are uniformly transported to the heating and mixing device 26 for secondary stirring. After the set stirring time is reached, visual judgment is performed again. If the electrolyte is still determined to be unqualified, it is sent to the sample storage area and the process is stopped, and it is stored as waste. The qualified electrolyte does not have sedimentation and layering, so the visual recognition device 27 determines whether the electrolyte is qualified by recognizing whether there is sedimentation and layering.

[0159] In this embodiment, the first mechanical arm 21 can be realized by a mechanical hand and a three-axis moving guide rail. The first mechanical arm 21 is controlled by the control module 10 to realize the functions of full automation of solid grinding, high-precision solid-liquid sampling, sample heating, sample shaking and mixing, and sample transfer.

[0160] In a feasible implementation, referring to Figure 4The test parameters include physical property parameters representing physical properties of the electrolyte and electrochemical parameters representing electrochemical properties of the electrolyte.

[0161] The electrolyte test module 30 includes a second mechanical arm 31, a plurality of dispensing containers 32, an electrochemical device assembly device 33, a physical property test device 34, and an electrochemical test device 35.

[0162] The control module 10 is configured to control the second mechanical arm 31 to dispense the configured electrolyte into the plurality of dispensing containers 32, to control the second mechanical arm 31 to send some of the dispensing containers 32 to the physical property test device 34 for physical property testing to obtain the physical property parameters of the electrolyte, and to control the second mechanical arm 31 to send some other of the dispensing containers 32 to the electrochemical device assembly device 33 for assembly to obtain electrochemical devices, and to control the electrochemical test device 34 to perform electrochemical testing on the electrochemical devices to obtain the electrochemical parameters of the electrolyte.

[0163] Specifically, as shown in FIG. 3, the control module 10 controls the electrolyte test module 30 to perform the following workflow for testing: Figure 4

[0164] (1) A conveying device 50 can be arranged between the electrolyte configuration module 20 and the electrolyte test module 30. The first mechanical arm 21 places the configured electrolyte reagent bottle 22 containing qualified electrolyte on the conveying device 50. The second mechanical arm 31 transfers the reagent bottle 22 on the conveying device 50 to an automatic dispensing area where a plurality of dispensing containers 32 are placed.

[0165] (2) The control module 10 controls the second mechanical arm 31 to dispense the electrolyte in the reagent bottle 22 into the plurality of dispensing containers 32 based on the input test method. The test method includes test items to be performed and the volume of electrolyte required for each test item. The number of test items corresponds to the number of dispensing containers 32 required for electrolyte dispensing, and the volume of electrolyte required for each test item corresponds to the volume of electrolyte to be injected into each dispensing container 32.

[0166] (3) In this embodiment, the electrolyte in the reagent bottle 22 is dispensed into the plurality of dispensing containers 32 for separate testing. This is to ensure the accuracy of the test results. The same portion of electrolyte is only tested once, and is not used for multiple tests.

[0167] ​(3) The second mechanical arm 31 sends the partial liquid container 32 filled with electrolyte to the physical property testing device 34 for physical property testing, and obtains the physical property parameters of the electrolyte. The physical property testing device 34 is integrated with a viscosity testing device, a Raman spectrometer, a nuclear magnetic resonance spectrometer, a saturated vapor pressure testing device, an electric conductivity temperature control testing device, etc. Among them, the viscosity testing device can test the viscosity of the electrolyte at different temperatures; the infrared spectrometer can test the functional groups of the electrolyte; the Raman spectrometer or the nuclear magnetic resonance spectrometer can test the solvation structure of the electrolyte; the saturated vapor pressure testing device can test the saturated vapor pressure value of the electrolyte at different temperatures; and the electric conductivity temperature control testing device can test the electric conductivity value of the electrolyte.

[0168] (4) The second mechanical arm 31 sends another partial liquid container 32 filled with electrolyte to the electrochemical device assembling device 33 for assembling, obtains the electrochemical device, and controls the electrochemical testing device 35 to perform electrochemical testing on the electrochemical device, and obtains the electrochemical parameters of the electrolyte.

[0169] The electrochemical testing device 35 is integrated with a high-throughput electrochemical workstation and a high-throughput charge-discharge tester, and the testing items include half-cell, full-cell coulomb efficiency testing, electrolyte limiting current density testing, electrochemical window testing, ion diffusion coefficient testing, and alternating current impedance testing, etc. Among them, the coulomb efficiency testing can measure the coulomb efficiency of the electrolyte, the limiting current density testing can measure the limiting current density of the electrolyte, the electrochemical window value testing can measure the electrochemical window value of the electrolyte, the ion diffusion coefficient testing can measure the ion diffusion coefficient of the electrolyte, and the alternating current impedance testing can measure the alternating current impedance of the electrolyte.

[0170] In the embodiment, the electrolyte needs to be assembled into an electrochemical device before electrochemical testing, such as a simple battery. For example, the electrolyte is filled into a simple device with a positive electrode and a negative electrode to assemble a simple battery.

[0171] In order to ensure the accuracy of the test results, the same electrochemical device can only be tested once, and multiple tests cannot be performed on the same electrochemical device during the electrochemical testing of the electrolyte. Therefore, the configured electrolyte needs to be assembled into multiple identical electrochemical devices, and different tests are performed on each electrochemical device.

[0172] The second mechanical arm 31 in the embodiment can be realized by a mechanical hand cooperating with a three-axis guide rail, and the first mechanical arm 21 and the second mechanical arm 31 can be the same or different.

[0173] The full-automatic equipment integrating multiple high-throughput tests in the embodiment realizes full-automatic testing by cooperation, and the test types cover the basic physical properties of various basic electrolytes, such as integrated novel automatic test methods of visual identification of electrolyte stratification / precipitation, conductivity test, saturated vapor pressure test, viscosity test, Raman spectrum test, nuclear magnetic resonance spectrum test, and the like; and integrated electrochemical test methods, such as half-cell, full-cell coulomb efficiency test, cyclic voltammetry test, limiting current density test, electrochemical window test, ion diffusion coefficient test, alternating current impedance test, and the like.

[0174] The present application is directed to the problems of long cycle time consumption, low efficiency, strong subjectivity of data analysis, and the like in the process of current manual development of electrolyte, configuration, test delivery, and analysis, and develops a full-automatic and intelligent liquid preparation and test characterization system, which can realize high-throughput automatic preparation and automatic characterization of electrolyte, and realize intelligent recommendation of electrolyte through AI technology. The present application can greatly improve experimental efficiency, realize precise control of experimental process in unmanned experimental operation, and complete repeated experiments autonomously through a robot, collect standardized experimental data, automatically realize high-throughput and standardized data whole-chain management, and finally provide synthesis optimization and scheme design guidance based on intelligent decision-making of artificial intelligence and data driving.

[0175] The above is only a specific implementation manner of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0176] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used for executing the x electrolyte recommendation method in the above-described embodiment.

[0177] The computer readable storage medium provided by the present application may, for example, be a U disk, but is not limited to an electric, magnetic, optical, electromagnetic, infrared, or semiconductor system, system, or device, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electric connection with one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present embodiment, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer readable storage medium can be transmitted in any suitable medium, including but not limited to an electric wire, an optical cable, an RF (Radio Frequency), etc., or any suitable combination of the above.

[0178] Computer program code for carrying out operations of the present application can be written in one or more programming languages or combinations of languages including object oriented programming languages such as Java, Smalltalk, C++ or conventional procedural programming languages such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0179] The flow and block diagrams in the drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flow and block diagrams can represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may be executed in the reverse order, depending on the functionality involved. It will also be noted that each block of the block diagrams and / or flowchart illustrations, and combinations thereof, can be implemented by special purpose hardware-based systems that perform the specified functions or operations, or combinations of special purpose hardware and computer instructions.

[0180] The modules involved in the embodiments of the present application can be implemented in software or in hardware. In some cases, the names of the modules do not limit the modules themselves.

[0181] The readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e., a computer program) for executing the above-mentioned electrolyte recommendation method, and can provide a new method that can improve the electrolyte formula development efficiency and development accuracy. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the electrolyte recommendation method provided by the above-mentioned embodiments, and will not be described here.

[0182] The present application also provides a computer program product comprising a computer program, which, when executed by a processor, implements the steps of the electrolyte recommendation method as described above.

[0183] The computer program product provided by the present application can provide a new method that can improve the electrolyte formula development efficiency and development accuracy. Compared with the prior art, the computer program product provided by the present application has the same beneficial effects as the electrolyte recommendation method provided by the above-mentioned embodiments, and will not be described here.

[0184] The above only describes some embodiments of the present application, and does not limit the patent scope of the present application. Any equivalent structural transformation made by using the content of the present application specification and drawings, or direct / indirect application in other related technical fields is included in the patent protection scope of the present application.

Claims

1. A method for recommending an electrolyte, characterized in that, The method includes: Multiple electrolytes are prepared based on multiple electrolyte formulations, wherein the electrolyte formulation includes multiple chemical components for preparing the electrolyte and the proportions of the multiple chemical components; Test parameters and theoretical parameters of each electrolyte formulation are obtained respectively. The test parameters are parameters characterizing the performance of the electrolyte obtained by testing the electrolyte, and the theoretical parameters are parameters characterizing the performance of the electrolyte obtained by theoretical calculation of the electrolyte formulation. Based on the various electrolyte formulations, the test parameters of the various electrolytes, and the theoretical parameters of the various electrolyte formulations, a pre-trained machine learning model is used to recommend electrolyte formulations that meet preset requirements. The machine learning model includes a correlation relationship, which characterizes the relationship between the electrolyte formulation and the test parameters and / or the theoretical parameters. The preset requirement includes that the test parameters and / or the theoretical parameters are within a preset range.

2. The method as described in claim 1, characterized in that, The associations mentioned include at least positive and negative correlations.

3. The method as described in claim 1 or 2, characterized in that, The pre-trained machine learning model is obtained in the following way: Collect sample data, which includes multiple sample electrolyte formulations, sample test parameters corresponding to the multiple sample electrolyte formulations, and sample theoretical parameters; The machine learning model is trained using a portion of the sample data to obtain the association relationship; The performance of the machine learning model is evaluated using another portion of the sample data that was not used in training; Based on the performance evaluation results of the machine learning model, the machine learning model is optimized until its performance meets the performance threshold.

4. The method according to any one of claims 1 to 3, characterized in that, The various electrolyte formulations include multiple electrolyte formulations composed of different types of chemical components in different proportions.

5. The method according to any one of claims 1 to 4, characterized in that, The test parameters include physical property parameters characterizing the physical properties of the electrolyte and electrochemical parameters characterizing the electrochemical performance of the electrolyte. The process of obtaining the test parameters for each electrolyte and the theoretical parameters for each electrolyte formulation includes: Each of the electrolytes was subjected to physical property tests and electrochemical tests to obtain the physical property parameters and electrochemical parameters of each electrolyte. Theoretical calculations were performed on each of the electrolyte formulations based on first-principles and / or molecular dynamics simulations to obtain the theoretical parameters for each electrolyte formulation.

6. The method as described in claim 5, characterized in that, The physical property tests include one or more of the following: viscosity test of the electrolyte at different temperatures, functional group test of the electrolyte, solvation structure test of the electrolyte, saturated gas pressure test of the electrolyte at different temperatures, and conductivity test of the electrolyte. The electrochemical tests include one or more of the following: coulombic efficiency test, limiting current density test, electrochemical window value test, ion diffusion coefficient test, and AC impedance test. The theoretical parameters include one or more of the following: density, ion binding energy, molecular dissociation energy, molecular surface electrostatic potential, dielectric constant, number of solvent donors, number of solvent acceptors, ion diffusion coefficient, and space charge density.

7. An electrolyte recommendation system, characterized in that, The system includes: a control module, an electrolyte preparation module and an electrolyte testing module connected to the control module, and a computing device connected to the electrolyte testing module; The control module is used to control the electrolyte preparation module to prepare multiple electrolytes based on multiple electrolyte formulations, wherein the electrolyte formulation includes multiple chemical components for preparing the electrolyte and the proportions of the multiple chemical components; It is also used to control the electrolyte testing module to test each of the configured electrolytes to obtain test parameters for each electrolyte, wherein the test parameters are parameters characterizing the performance of the electrolyte obtained by testing the electrolyte; The computing device is used to acquire test parameters for each electrolyte; it is also used to acquire theoretical parameters for each electrolyte formulation, wherein the theoretical parameters are parameters characterizing the performance of the electrolyte obtained by theoretical calculation of the electrolyte formulation; It is also used to recommend electrolyte formulations that meet preset requirements based on the multiple electrolyte formulations, the test parameters of the multiple electrolytes, and the theoretical parameters of the multiple electrolyte formulations, using a pre-trained machine learning model; wherein, the machine learning model includes correlation relationships, which characterize the relationship between the electrolyte formulation and the test parameters and / or the theoretical parameters, and the preset requirements include: the test parameters and / or the theoretical parameters are within a preset range.

8. The system as described in claim 7, characterized in that, The electrolyte preparation module includes: a first robotic arm, a reagent bottle, a powder injection device, a liquid injection device, a weighing balance, a heating and mixing device, and a visual recognition device; The control module is used to control the first robotic arm to pick up the reagent bottle and place the reagent bottle on the weighing balance; The powder injection device is used to inject powder into the reagent bottle on the weighing balance based on one of the multiple electrolyte formulations. The liquid injection device is used to inject solvent into the reagent bottle on the weighing balance based on the electrolyte preparation direction; The control module is also used to control the first robotic arm to place the reagent bottle on the weighing balance into the heating and mixing device after the powder injection device and the liquid injection device have completed their operations. The heating and mixing device is used to heat and mix the powder and solvent in the reagent bottle. The control module is also used to control the first robotic arm to place the heated and mixed reagent bottle onto the visual recognition device. The visual recognition device is used to identify whether the liquid in the reagent bottle is qualified; The control module is also used to control the first robotic arm to remove the qualified reagent bottle and obtain the prepared electrolyte.

9. The system as described in claim 7, characterized in that, The test parameters include physical property parameters characterizing the physical properties of the electrolyte and electrochemical parameters characterizing the electrochemical performance of the electrolyte. The electrolyte testing module includes: a second robotic arm, multiple dispensing containers, electrochemical device assembly equipment, physical property testing equipment, and electrochemical testing equipment; The control module is used to control the second robotic arm to dispense the prepared electrolyte into the plurality of dispensing containers; It is also used to control the second robotic arm to send part of the liquid separation container into the physical property testing equipment for physical property testing, so as to obtain the physical property parameters of the electrolyte; It is also used to control the second robotic arm to send another part of the liquid separation container into the electrochemical device assembly equipment for assembly to obtain an electrochemical device, and to control the electrochemical testing equipment to perform electrochemical testing on the electrochemical device to obtain the electrochemical parameters of the electrolyte.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the electrolyte recommendation method as described in any one of claims 1 to 6.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the electrolyte recommendation method as described in any one of claims 1 to 6.

Images